OurAirports Data

It is the reference layer the rest of aviation data stands on. Where traffic, delay and fare datasets measure what happened, OurAirports Data supplies who and where: roughly 85,900 airports across every ISO country, each pinned by ICAO, IATA, GPS and local codes, decimal-degree coordinates, elevation, continent, country and region codes, the municipality served, a seven-value facility type and a scheduled-service flag.

The project started in 2007, built by volunteers after public access to the DAFIF aviation dataset was withdrawn in 2006, and it has been filling that hole ever since - which is why flight-planning apps, mapping services and humanitarian platforms treat it as the registry of record. Around each master row sit its companions: runways with dimensions, surface, lighting and closure status, the voice frequencies serving the field, and the radio navaids pointing at it.

Get a sample of this dataset and see your market's airports already resolved.

Three master rows and two runway rows, exactly as the fields arrive:

# AIRPORT - one row per facility
id          : 6523          ident : 00A         type : heliport
name        : Total RF Heliport
municipality: Bensalem      region: US-PA       country: US
lat/lon     : 40.070985, -74.933689              elev_ft: 11   scheduled_service: no

# AIRPORT
id          : 323361        ident : 00AA        type : small_airport
name        : Aero B Ranch Airport
municipality: Leoti         region: US-KS       country: US
lat/lon     : 38.704022, -101.473911             elev_ft: 3435 scheduled_service: no

# AIRPORT
id          : 6524          ident : 00AK        type : small_airport
name        : Lowell Field
municipality: Anchor Point  region: US-AK       country: US
lat/lon     : 59.947733, -151.692524             elev_ft: 450  scheduled_service: no

# RUNWAY (joined on airport_ref)
airport     : 00A (6523)    size  : 80 x 80 ft  surface: ASPH-G
lighted     : 1             closed: 0           low end: H1

# RUNWAY
airport     : 00AK (6524)   size  : 2500 x 40 ft  surface: GRVL
lighted     : 0             closed: 0   low end: N   high end: S

Two things read straight off those lines. First, the spread: a Pennsylvania hospital heliport with an 80-foot pad, a Kansas ranch strip at 3,435 feet, an Alaska gravel runway - none of which appear in hub-centric lists. Second, the join: each runway keys on airport_ref, so one airport query returns its entire operating picture.

Twenty-seven fields form the confirmed core across the four record families - airport master, runway, frequency, navaid - each verified against the publisher's own data dictionary rather than guessed from headers. The id column is the spine, and the four code columns split honestly by airport tier.

Geography - worldwide: roughly 85,900 facilities across every ISO country, tagged by continent code AF through SA, with administrative subdivisions carried beneath the country level so regional rollups come without a crosswalk.

Temporal - this is a current-state registry: one authoritative snapshot of each facility as it stands today, catalogued continuously since 2007 and published as latest state rather than a year-by-year series. Treat it as the definitive now of world aviation infrastructure; openings, closures and runway changes over time need a historical counterpart paired in.

Granularity - one master row per airport, joined one-to-many to its runways, voice communication frequencies and radio navigation aids. Set against the wider catalog - average quality score 7.81 across all 1,744 datasets Datadory catalogs - this slice scores 9/10, verified end to end with sample rows confirmed.

API, files, or your warehouse. Daily, weekly, or hourly.

Your cadence is decoupled from anything upstream - take one full snapshot of the registry, or keep the warehouse current so new facilities and changed runways diff cleanly against last week's rows. Every delivery arrives keyed on the persistent id and normalized to the field dictionary above, with the loose vocabulary columns flagged rather than discovered. Sample first: name the countries, regions or facility types and the extract lands cut to exactly that scope.

• Geocoding and enrichment - turn the bare codes inside accident, traffic and route feeds into named, geocoded places with region and country attached, so downstream models stop joining on strings alone.
• Product features - autocomplete, map pins and nearby-airport logic drawn from 85,900 typed rows rather than a hand-maintained list.
• Reachability studies - runway length, width, surface and lighting decide whether a given aircraft can accept a given destination.
• Market sizing - airports and runways counted by country, region and facility type expose coverage gaps and market opportunity.
• Competitive benchmarks - measure any mapping or flight product's airport coverage against the same master list.
• Citation-grade sourcing - journalists, academics and humanitarian platforms anchor claims to the neutral registry of record.

Developers and builders rank first: autocomplete, pins and scheduling features ship faster on a typed, keyed reference layer that survives code reassignments; see developers builders use cases. Data scientists and ML engineers get the join layer underneath every code-keyed aviation feed, with a persistent id making the merge reproducible; see data scientists use cases. Market researchers and consultants count real infrastructure by geography instead of approximating from hub lists. Competitive-intel teams (their use cases) benchmark rivals' coverage against the same 85,900 records, and journalists, academics and students cite a registry that includes the balloonports.

Within aerospace and defense data, this record owns the world airport master: every ISO country, every facility class from large_airport down to balloonport, one uniform schema. The neighbours own different jobs. The NTSB Aviation Accident & Incident Database measures what went wrong at particular places and is the natural pairing - events gain coordinates and region context when joined through this layer (the head-to-head). The FAA Aircraft Registry covers US aircraft ownership and airworthiness, not places. The OpenFlights record holds a useful but frozen network layer - airlines, routes and a smaller airport list - while EUROCONTROL's Aviation Data & Dashboard measures European performance rather than cataloguing world facilities. If the question is where can aircraft go, everywhere, this is the record.

Three things worth knowing upfront. First, the code columns split by airport tier: icao_code and iata_code populate at the larger fields and thin out quickly below them, while gps_code and local_code carry the small-airfield load - pipelines that assume one universal code will drop thousands of legitimate rows. Second, two vocabularies are documented but deliberately loose: runway surface and frequency roles arrive as codes rather than a fixed enum, so group-by queries should expect variety. Third, this is a snapshot, not a historical series - anyone needing openings, closures or runway changes over time should say so, and the historical counterpart gets paired in rather than the snapshot oversold. Name your countries, regions or facility types and the sample returns in exactly the schema shown above.